As a rat navigates through space, neurons located throughout the limbic system provide an abstract representation of the environment. The cells which provide this map-like representation fall into two general categories. One type, known as Place cells, fires optimally whenever the rat is in one particular location within the larger environment. Each Place cell has its own region of high firing, so that for each spot the animal visits, there is a unique set of cells which will be active. The other type is known as Head Direction cells. Each of these fires whenever the rat faces one particular direction and will do so regardless of the rat's current location or behavioral state. Since each cell has its own preferred direction, there is a unique set of directional cells which are active for each possible heading. Previous work; has shown that both the Place cells and the Head Direction cells rely, in part, on a process known as path integration, to somehow "track" the rat's location and directional heading, respectively. Path integration involves the use of the animal's own movement to constantly update the spatial firing patterns. For example, if the rat begins at, location 'A', and then takes a few steps to arrive at 'B', this motion itself will, somehow, cause the 'A' Place cells to turn off, and the 'B' Place cells to turn on. Similarly, if the animal begins with activity in the 'north'-signaling Head Direction cells, and then turns 90 degrees to the right, this angular motion will, itself, somehow turn the 'north' Head Direction cells off, and the 'east' Head Direction cells on. Recent data suggest that the Place and Head Direction systems may each have their own path integration circuitry located within particular, identified brainstem components of the limbic system. The goal of the proposed work will be 1) to record from cells in these nuclei to determine whether they contain signals which have been postulated as necessary for path integration in neural network models of Place and Head Direction cells, and 2) to lesion these areas and examine the effects on Place and Head Direction cells in other parts of the limbic system. We postulate that lesions of these nuclei should destroy all Place and Directional signals throughout the limbic system. If it is verified that these nuclei do, in fact, constitute the postulated path integration circuitry, this will provide a basis for important future work. Path integration is a sophistocated neural computation thought to underly complex cognitive mapping abilities. In addition, the circuitry postulated for path integration is formally similar to that postulated for numerous other cognitive functions, including short term memory and certain types of visual perception. Thus the identification of the path integration circuitry for the Place and Head Direction systems would allow for detailed study of a type of circuit which is likely to play a central role in several types of higher cognitive function.